The Predictive Analyst
Spring
2017
Spotlight on Hidden Physicists
The Predictive Analyst
Meghan Anzelc, Lead Data Scientist for Zurich North America
I actually came late to physics. I didn’t take it in high school, but in college a professor encouraged me to take calculus-based physics to really learn it. It turned out that course was only for physics majors, but I took the course, thinking I’d change my major after a year.
That summer I did a research internship in atmospheric chemistry, studying trace metals in aerosols, so I stuck with physics for a second year. I did another research internship the following summer, and that was it. I was committed to pursuing a physics degree. I loved that you could design an experiment, collect data and analyze them, and learn something about the way the world worked that was unknown before. That carried me into pursuing a doctorate.
Fast forward a few years. As I was doing my dissertation work at Fermilab, I realized that there were very few pure research jobs anymore, and for much of the time I was there, a hiring freeze was in place for lab employees. I saw a lot of people around me who were on their second, third, or fourth post-doc position. And as with many things, as you get into the reality of what a job is like, you realize it’s different than what you thought it would be, in both good and not so good ways.
I started informational interviewing, talking to contacts I knew from friends and using my university alumni directory and the APS directory. Finally, a friend of mine connected me to someone working in predictive modeling at an insurance company. I reached out to do an informational interview, and at the end of the conversation he said his team was hiring and encouraged me to apply. This led to interviews and a job offer I accepted.
Now I lead a team of over 40 data scientists, all focused on using predictive analytics to solve business problems in insurance. For example, we use analytics to help better understand our customers’ risks, assessing and quantifying them to appropriately price insurance coverage for them. We also use analytics to help ensure the right information is provided to the people who interact with customers, so they can help customers more quickly and effectively.
This is a growing field and everyone sees opportunities to use predictive analytics for their team—which is exciting!—but because of timing or costs we can’t take on every request. That can be frustrating, for sure, but it shows how exciting this field is right now.
The most challenging part of my job is probably bridging the gaps between teams. Because of my background and training, I tend to see analytics as the solution to a problem. Someone in a different department may see the problem differently. It doesn’t mean one of us is wrong and the other is right, but it does mean that we have to spend time to really understand each other’s viewpoints as well as the specifics of the problem we’re trying to solve. The challenge can be more difficult if we don’t make the effort to really try to understand each other. However, I learn so much from hearing about viewpoints that are different from my own, and I really believe in the end we come up with better solutions by working together.
Another thing I get to do in my job is provide development opportunities for everyone on my team, whether through a project assignment where a team member can learn a new tool, or by coaching a manager through a tough situation. It’s rewarding to watch people on my team successfully take on new challenges and grow in their careers. I love helping people with their careers and offering advice based on what I’ve learned, which I’ve tried to do since graduate school. Here’s one piece of advice: Spend time thinking about what you really enjoy, what you don’t, and where your strengths lie. This can help you identify a wider range of career paths than perhaps you have already considered, as a lot of skills are transferable and valuable to companies.
Of course, I think there is a lot of exciting work going on at Zurich and think you should consider joining us, too! There are lots of untapped opportunities to use data to solve problems, and it’s exciting to be part of that conversation and shaping what we do next.